Accuracy validation study

Accuracy is the closeness of agreement between the value a method reports and the value accepted as true. ICH Q2(R2) asks for it to be demonstrated across the method's range, with a minimum of nine determinations over at least three concentration levels — three replicates at each — and reported as percent recovery of a known added amount, together with the confidence interval.

This study groups your determinations by the amount added, computes each level's mean recovery, its relative standard deviation and a two-sided confidence interval from the tabulated Student's t, and a grand mean weighted by how many determinations each level carries. Every value is stored in ppm; you enter in the unit you work in and the conversion happens once, at the edge, so the record never carries a unit ambiguity.

Under ICH Q2(R2), USP <1225> and USP <1226> the conventional default is a mean recovery of 98–102 % at every level — an assay-style figure that the guideline leaves to the study protocol, and that a lab working in a difficult matrix will set wider. Under AOAC the criterion is the Appendix F expected range at each level's concentration: 80–110 % at 10 ppm, widening as the analyte gets scarcer. The band is looked up at the level's expected concentration — native plus added — so a small spike on a large native background is judged at the right place.

The amounts added are concentration increments, correct whenever the spike volume is negligible against the sample. A spike large enough to dilute the sample needs the volume-corrected form, which the free spike recovery calculator provides for a single determination.